Professional Documents
Culture Documents
By:
Abhineet Kumar Sharma
Roll No. 80681001
JUNE 2008
MECHANICAL ENGINEERING DEPARTMENT
THAPAR UNIVERSITY
PATIALA 147004
ACKNOWLEDGMENT
Working without proper guidance and expecting success is just like making castles in
the air, so whenever one wants to start any work, he requires guidance from experts. I
express my sincere gratitude to my guide ,Mr. Satish Kumar, lecturer , Department of
Mechanical Engineering, Thapar University, for acting as supervisor and giving
valuable guidance during the course of this investigation, for his ever encouraging and
timely moral support. His enormous knowledge and intelligence always helped me
unconditionally to solve various problems.
I am greatly thankful to Dr. S. K. Mohapatra, Professor and Head, Mechanical
Engineering Department, Thapar University for his encouragement and inspiration
for execution of the thesis work. I do not find enough words with which I can
express my feeling of thanks to entire faculty and staff of Department of Mechanical
Engineering, Thapar University, Patiala for their help, inspiration and moral support,
which went a long way in successfully completion of my thesis.
ii
ABSTRACT
Slurry erosion is a crucial problem in many industrial applications, due to
impingement of solid particles being suspended in the fluids flowing at high velocity.
It affects both the initial cost and life of the component. Erosive wear can be found on
impellers & volute casing in slurry pumps, angled pipe bends, turbines, pipes and pipe
fitting, nozzles, burners etc. The volume loss due to erosion is a troublesome problem
for slurry transportation. The service life of equipments handling solidliquid mixture
can be increased by reducing the erosion wear. It is therefore necessary to study
erosion wear systematically, so that efforts can be made to minimize the wear during
design stage. Centrifugal pumps are generally used for slurry transport. Due to
erosion, the life of pumps used is very short and they need to replace periodically.
In the present work a systematic study on the erosion wear has been carried out to
improve the present understanding .Numerical simulation has been carried out for
radial flow impeller and the effect of affecting parameters on erosion wear has been
studied .The results clearly indicated that all factors contribute to wear of centrifugal
pump impeller. The erosion rate of blades is 10 times greater than the erosion wear of
hub. The erosion rate is found to be more severe near the leading edge as compared to
other parts of impeller. The simulations which were carried out on impeller show that
particle size affects the erosion rate only up to certain size of particles and depends on
the impact angle.
iii
CONTENTS
TITLE
PAGE No.
Certificate
Acknowledgement
ii
Abstract
iii
Contents
iv
List of Figures
vii
List of Tables
xi
Nomenclature
xii
CHAPTER 1: INTRODUCTION
(1-18)
1.1 SLURRY
1.3. PUMPS
1.4.2 Advantages
11
iv
1.5 WEAR
11
11
14
15
16
16
17
18
18
19-27
28-36
28
28
37-51
FLUID DYANAMICS
4.1 INTRODUCTION TO CFD
37
38
38
38
38
39
4.3.4 Pre-processing
42
4.3.5 Solver
50
v
51
53-57
53
53
53
54
5.5 PREPROCESING
56
55
56
56
5.6 SOLVER
CHAPTER 6: RESULTS AND DISCUSSIONS
58
59-71
6 .1 DESIGN OF IMPELLER
59
60
OF IMPELLER
6.3 WEAR PATTERN ON THE HUB
62
63
ON EROSION RATE
6.5 EFFECT OF FLOW RATE
65
ON EROSION RATE
6.6 EFFECT OF SPEED ON THE EROSION RATE
65
67
68
68
ACROSS BLADES
CHAPTER 7: CONCLUSIONS AND SCOPE OF FUTURE WORK 72-73
7.1 CONCLUSIONS
72
73
REFERENCES
74-77
78-82
CENTRIFUGAL PUMP
ANNEXURE II: OUTPUT OF THE PROGRAM
85-86
vii
LIST OF FIGURES
Figure No.
Description
Page No.
1.1:
1.2
Reciprocating pump
1.3
Rotary pump
1.4
1.5
1.6a
1.6b
1.7
Open impeller
1.8
Semi-open impeller
1.9
Closed impeller
1.10
10
1.11
16
1.12
17
3.1
29
4.1
39
4.2
41
4.3
41
4.4
41
4.5
41
5.1
53
5.2
54
5.3
55
6.1
60
6.2
61
viii
6.3
61
6.4
62
62
6.6
63
6.7
64
6.8
64
65
on blade
6.10
66
6.11
66
6.12
67
67
6.14
69
6.15
69
6.16
70
6.17
70
ix
LIST OF TABLES
Table No.
Description
Page No.
3.1
36
5.1
56
5.2
57
58
NOMENCLATURE
SYMBOLS
STANDS FOR
Discharge in m3 / sec
Net head in m.
Ns
Specific Speed
Po
Output power
Pin,
Input power
Psh
Shaft Power
Hydraulic Efficiency
Overall Efficiency
Mechanical Efficiency
Fs
Dsh
Shaft diameter
Dhb
Hub diameter
CA
C1
C3
D1
b1
u2
u1
Hr
Number of blades
C2m
C1m
B2
r2
Fs
Shear stress
eff
Effective
viscosity
accounting
for
turbulence
Turbulence viscosity
Turbulence frequency
F2
Blending function
Strain rate.
Continuous phase
Dispersed phase
Vp
v pi
Particle velocity
Time in seconds
Vp
f ( )
B1
Pfleiderers coefficient
Turbulent viscosity
xiii
CHAPTER 1
INTRODUCTION
______________________________________________
1.1 SLURRY
Slurry is a mixture of solids and liquids. Its physical characteristics are dependent on
many factors such as size and distribution of particles, concentration of solids in the
liquid phase, size of the conduit, level of turbulence, temperature, and absolute (or
dynamic) viscosity of the carrier. Nature offers examples of slurry flows such as
seasonal floods that carry silt and gravel. A slurry mixture is a mixture of a carrying
fluid and solid particles held in suspension. The most commonly used fluid is water,
however in some cases air is also used such as in pneumatic conveying. Theoretically,
a single-phase liquid of low absolute (or dynamic) viscosity can be allowed to flow at
slow speeds from a laminar flow to a turbulent flow. However, a two-phase mixture,
such as slurry, must overcome a deposition critical velocity or a viscous transition
critical velocity. If the slurrys speed of flow is not sufficiently high, the particles will
not be maintained in suspension. On the other hand, in the case of highly viscous
mixtures, if the shear rate in the pipeline is excessively low, the mixture will be too
viscous and will resist flow.
may form in the pipe, with the heavier particles at the bottom and the lighter ones in
suspension. Heterogeneous slurries are encountered in many places mining, phosphate
rock mining, and dredging applications. Heterogeneous flows require a minimum
carrier velocity.
Although the transportation of solids in the form of slurries is basically older than
history.The blood circulating system in mammals involves the use of a positive
displacement pump forcing slurry of solid corpuscles in liquid serum through a
complex pipeline. The slurry transportation of solids through long pipelines and with
the help of centrifugal pumps has been undertaken only in almost last 25 years.
1.3 PUMPS
A pump is a device used to move liquids or slurries. A pump moves liquids from
lower pressure to higher pressure. Machine that draws a fluid into itself through an
.
Figure 1.3: Sectional view of rotary pump
3
.
Figure 1.4: Model of a centrifugal pump
pumped by forcing the fluid outwards from the centre of rotation. Figure 1.6a shows
the axial, radial and tangential component of flow.
Components of impeller
a) Blade
Blades are the series of backward or forward curved vanes which transfers the power
from shaft to the fluid.
b) Hub and Shroud
The hub is the surface of the machine closest to the axis of rotation. It defines the
inner fluid flow surface. The shroud is the surface of the machine farthest from the
axis of rotation. It defines the outer fluid flow surface. The hub and shroud can be
defined only after the machine data has been defined, although all of these objects can
be defined in one step.
c) Leading and trailing edges
The leading edge curve is the most upstream part of the blade. Any change to the
leading edge changes the blade surfaces, which changes the periodic surfaces as well
as the hub and shroud surfaces. The trailing edge curve is the most downstream part
of the blade.
c) Enclosed impeller
Totally enclosed impellers are designed with the blades between two support shrouds
or plates. These impellers are generally used clean liquids because tolerances are tight
at the eye and the housing,and there is no room for suspended solids, crystals or
sediment, Figure 1.9. shows a type of enclosed impeller
b) Vortex casing
If a circular chamber is introduced between the casing and the impeller, the casing is
known as vortex casing. By introducing the circular chamber, the loss of energy due
to formation of eddies is reduced considerably. Thus, the efficiency of the pump is
more than the pump with volute casing.
c) Casing with guide blades
In this type of casing, the impeller is surrounded by series of guide blades mounted on
a ring, which is known as diffuser. The guide vanes are designed in such a way that
water from the impeller enters the guide vanes without stock. Also the area of guide
vanes increase, thus reducing the velocity of flow through guide vanes and
consequently increasing the pressure of pressure of water. The water from the guide
vanes then passes through the surrounding casing, which is concentric with the
impeller
10
4) Delivery pipe
A pipe whose one end is connected to the outlet of the pump and the other delivers
the water at the required height is known as delivery pipe.
1.4.5 Centrifugal pump applications
Pumps are used wherever any quantity of liquid must be moved from one place to
another. Pumps are found in such services as steam power plants; water supply plants;
sewage; drainage or irrigation; oil refineries, chemical plants and steel mills; food
processing factories and mines; dredging or jetting operations; hydraulic power
services and almost every ship whether driven by diesel or steam engine. While these
pumps have much in common, they are varied to meet special requirements and
particular needs of each service.
Petroleum Industry
Chemical Industry
Textile Industries
Paper Industry
Sewage and Sump Services
Irrigation, Drainage and Flood Control
Mining and Construction
1.5 WEAR
Wear is defined as progressive volume loss of material from a solid surface due to
corrosion, abrasion and erosion. Wear is one of the most common problems
encountered in industrial applications. It is defined as the erosion of material from a
solid surface by the action of another solid. In the domain of wear, particularly the
wear encountered in handling abrasive solid particles, much work has been done in
the past half century with regard to dry abrasivity, but only in more recent years has
interest grown in wet abrasivity, namely slurries. Wear can be classified into
following types
11
reaction between metal and oxygen. These oxides are wiped away with the flow and
cause pitting of the surfaces. Corrosion is accelerated as impacted surfaces are
exposed to slurry chemistry.
1.5.1.4 Surface fatigue wear
Surface fatigue is a process by which the surface of a material is weakened by cyclic
loading, which is one type of general material fatigue. Surface fatigue wear occurs as
result of the formation and growth of cracks. It is the main form of wear of rolling
devices such as ball bearings, wheels on rails, and gears. During continued rolling, a
crack forms at or just below the surface and gradually grows until a large particle is
lifted right out of the surface.
1.5.1.5 Erosive wear
Erosive wear is the dominant process and can be defined as the removal of material
from a solid surface due to mechanical interaction between the surface and the
impinging particles in a liquid stream. Erosion involves the transfer of kinetic energy
to the surface. This means that in erosion material removal is a function of particle
velocity squared to higher power. Erosive wear depends on the predominant impact
angle of particle impingement with the material surface. Impact angle will vary from
0 to 90 degrees and depend on both fluid particle and particle- particle interaction.
This type of wear can be found on impellers and volute casing in slurry pumps, angled
pipe bends, turbines, pipes and pipe fitting, nozzles, burners etc. The material loss due
to erosion increases with the increase in kinetic energy of the particles impacting at
the target surface.
The volume loss due to erosion is a troublesome problem foe slurry transportation
systems e.g. mineral transport systems, ash disposal systems. The erosion wear due to
the air borne particles in some devices such as jet planes and turbines is also
significant due to very high impact velocity. It is thus a challenging task to control the
erosion wear in many engineering applications. The material removal due to erosion
is caused by two dominant mechanisms namely brittle fractures and platelet
deformations.
13
In brittle type material, the solid particles impacting on the target surface forms cracks
in longitudinal and lateral directions. These cracks propagate due to impact of
succeeding particles and broken materials pieces will be carried out by flowing fluid.
The material removal rate due to brittle fracture increases with increase in normal
component if the particle velocity and thus the brittle type material show maximum
wear near normal impact angles.
In Platelet mechanism, the impact of solid particles deforms the target surface to
forms hills and valleys. The repeated impacts of particles remove the material and
forms crater at the surface. This mechanism along with micro-cutting and chipping
dominates in ductile type materials, which show the maximum wear in the impact
angle range of 20-40 degree. Apart from the target surface characteristics like brittle
or ductile type, many other parameters such as solid particles, carrier fluid, flow
conditions etc. affect the erosion wear. It is, therefore, difficult to estimate wear for a
given operating conditions.
1.5.2 Parameters affecting erosion wear:
The prominent parameters and their effect on erosion wear are as under:
1) Impact angle
Impact angle is defined as the angle between the target surface and the direction
striking velocity of the solid particle. The variation of erosion wear with the impact
angle depends on the characteristics of the target surface material namely brittle or
ductile type.
2) Velocity of solid particles
Velocity of solid particle strongly affects the erosion wear. As particle velocity
increases there is significant increase in erosion rate. The erosion rate is generally
related to the particle velocity using power law relationship in which the power index
for velocity varies in the range of 2-4.
3) Hardness
Hardness is the characteristic of a solid material expressing its resistance to permanent
deformation.Surface hardness as well as hardness of solid particles has profound
effect on the erosion wear mechanism. Hardness ratio has been defined as the ratio of
hardness of target material to the hardness of solid particles.
14
Figure 1.11 (a) Jet erosion tester (b) Slurry pot test (c) Coriolis erosion test
1.6.2 Slurry pot test
In slurry pots, a specimen rod of circular cross-section, fixed to the end of an arm, is
rotated in a circular container filled with the slurry assuming homogeneous
suspension of the solid particles. In slurry pot tests, also the amount of material
removed is determined by the weight loss. The samples are weighted before and after
the tests.
16
Slurry pot tester is simple in design, easy to fabricate and operate. In a slurry pot
tester, generally two cylinder wear specimens have been rotated in solid liquid
mixture. The rotational movement of the wear specimens and sometimes an impeller
attached at the end of 6th shaft keep the solid particles suspended in the liquid. The
rotating test specimen moves at a velocity relative to the solid-liquid suspension,
which is assumed stationary and homogeneous inside the pot.
In a slurry pot tester the rotational velocity of the specimen has been taken as the
relative velocity between the particles and wear specimens assuming suspension as
stationary. However, due to rotation of the wear specimen and propeller the mixture
inside the pot are highly turbulent conditions. Baffles are used to break the vortex
produced, which also develops some amount of non-homogeneity in the suspension.
These phenomena result in some error in evaluating the independent/isolated effect of
various parameters such as velocity, concentration, and impact angle etc on erosion
wear.
1.6.3 Coriolis erosion test
direction of rotation and is eroded by slurry particles, which are pressed against the
specimen face by the Coriolis acceleration. The erosion groove is transversed using a
profilometer and the local erosion depth is determined to with in a few microns. In the
Coriolis erosion tester, a relatively small batch of slurry, from an over-head tank
passes through the rotor in one pass. At rotor speeds of 6000 rpm, a measurable
groove is worn into the specimen within a few minutes. Figure 1.12 shoes the
schematic diagram of Coriolis erosion testing
1.6.4 Rotating pipe test
It consists of a circular pipe with wear pieces fitted at its inside surfaces and is filled
with slurry with half of its capacity. The pipe ring was rotated at its own centre and
relative motion between slurry and wear specimen causes erosion wear. This method
is used to simulate pipeline erosion, but takes long time to generate wear data.
1.6.5 Concentric cylinder viscometer
In this type of test one cylinder was rotated inside a cylinder containing slurry. The
wear at the outer wall of the inner cylinder surface was assumed to be cause by
parallel flow of slurry.
18
CHAPTER 2
LITERATURE REVIEW
___________________________________________________
Desai P.V1 [1990] has observed that erosion wear of centrifugal slurry pumps is
primarily governed by the particulate motion and concentration as well as their
physical properties. The analysis and the finite element computations yield the solid
velocity and concentration fields in an arbitrary radial cross-section of a centrifugal
slurry pump casing. The solutions were examined in light of their applicability to the
pump wear problem. Axisymmetric finite elements have to be used to analyze the
flow in the volume of revolution. The shape factor of the particles is introduced into
the drag and pressure force calculations to account for the angularity of the particles.
Achim Daniela et al2 [1999] have studied the computational and experimental data of
tube erosion in a fluidised bed and simulated the work with CFX code having
computational model of hydrodynamics and Finnie model of erosion. The results
predicted that most of wear occurs around the bottom of tube at an oblique impaction
angle and very low value at the top of the tube.
Manickam M. et al3 [1999] have modelled the flow in the bifurcation duct of a
power generation boiler plant. The computational fluid dynamics code was
customized to determine erosion rate caused by particles that hit the duct wall. To
reduce the wall duct erosion, five modifications to the duct geometry were tested
using the CFD model. The focus of these modifications have been used to improve
particle flow characteristics to mitigate the erosion of the walls of the duct by using
extended trailing edged or setting up more baffles in the central zone where the
particle traffic is intense.
Anders Sellgren et al5[1999] have studied the effects of non-Newtonian mineral
suspensions on the performance of centrifugal pumps. They have investigated how the
nonNewtonian rheology of some industrial suspension influences the performance of
centrifugal pumps. Slurry used in this experiment is red mud associated with
production of aluminium. It is observed that influence on the performance is normally
small with the values of kinematic viscosity of fluids less than 0.1 10 3 . Also
19
reductions in pump head and efficiency in the best efficiency point (BEP) region was
at a maximum about 10 and 20%. It has also been observed that pumping highly
viscous fluids can cause large drop in efficiency.
Lee S. Y. et al6 [2001] have discussed the application of computational fluid
dynamics (CFD) methods to simulate the flow of slurry and predict the erosion rates
so that an effective maintenance schedule can be developed for the filtration system of
the waste treatment process. The location of the maximum erosion for the selected
components is also identified. Solids content of the working fluid, the regions of high
wall shear, and particle impingement with the walls were considered as major
mechanisms associated with the erosion. All these tests were performed using sandwater slurry.
Gandhi B. K. et al7 [2001] have observed that performance of pumps decreases for
increase in solid concentration, particle size and specific gravity. The head and
efficiency of the pump decrease with increase in solid concentration, particle size, and
slurry viscosity, the decrease in the head being 210 percent higher than that of the
efficiency. The presence of finer particles (less then 18micron) in coarser slurries
substantially attenuates the loss of performance of the pump in terms of head and
efficiency. At low solid concentrations less than 30 percent by weight, the increase in
the pump input power is directly proportional to the specific gravity of slurry whereas
the same relationship is not applicable at higher concentrations. The study on the
pumps has confirmed that the additional head loss for slurries decrease with increase
in the pump size.
Hawthorne H.M.8 [2001] has conducted Coriolis tests for the evaluation of slurry
erosion on different materials. Slurries consisting of glass beads of size 90- 200micron size with 10% slurry concentration were taken and tests were performed on
1020 steel and copper at different impingement angles of 90-20 degrees. It was also
observed that in slurry jet testing, most particles impact the specimen above its critical
velocity resulting severe plastic deformation. In contrast, in the Coriolis test most
particle impacts result in only elastic deformation or mild plastic deformation. Hence,
elastic as well as plastic properties of specimen materials affect their performance in a
Coriolis slurry erosion evaluation, thus the results obtained from Coriolis tests were
more accurate.
20
Clark H. McI.10 [2002] has investigated the effect of Particle velocity and particle
size in slurry erosion. A list of factors affecting slurry erosion such as concentration
of particles, Slurry flow speed (particle impact speed), particle impact angle ,particle
size, particle density, hardness, friability, nature of suspending liquid, nature of slurry
flow (esp. local turbulence),nature of target material were explained. It is emphasized
that material loss must be measured by changes in surface profile rather than mass
loss, and that the best specimen form for this analysis is a cylinder.
Zhou Weidong et al11 [2003] have investigated the flow through a centrifugal pump
impellers using computational fluid dynamics and describes the three-dimensional
simulation of internal flow in three different types of centrifugal pumps A commercial
three-dimensional CFX code, with a standard two-equation turbulence model was
used to simulate the problem under examination . In the calculation, the finitevolume method and an unstructured grid system were used for the solution
.Comparison of computational results for various types of pumps showed good
agreement for the twisted-blade pumps.
Hamed Awatef A. et al12 [2002] have observed the blade surface deterioration by
erosion. A combined experimental and computational work was to investigate turbine
vane and blade material surface deterioration caused by solid particle impacts. Tests
were conducted in the erosion wind tunnel for coated and uncoated blade materials at
various impact conditions. Surface roughness measurements obtained prior and
subsequent to the erosion tests were used to characterize the change in roughness
caused by erosion. Numerical simulations for the three dimensional flow field and
particle trajectories through a low pressure gas turbine were employed to determine
the particle impact conditions with stator vanes and rotor blades using experimentallybased particle restitution models.
Engin Tahsin et al13 [2003] have evaluated some existing correlations to predict head
degradation of centrifugal slurry pumps. A new correlation has been developed in
order to predict head reductions of centrifugal pumps when handling slurries. The
proposed correlation takes into account the individual effects of particle. The
proposed correlation is therefore recommended for the prediction of performance
factors of small-sized slurry pumps having impeller diameters lower than 850 mm.
21
size, particle size distribution, specific gravity and concentration of solids, and
impeller exit diameter on the pump performance.
Clark H.McI14 [2004] has studied the influence of the squeeze film in slurry erosion.
A squeeze film may be taken as any liquid layer separating two approaching
surfaces. The presence of a squeeze film generally leads to a significant retardation of
any particle closely approaching a wearing surface at any speed and may even prevent
direct impact altogether. The experiments were performed in slurry pot tester. In
addition to this it is has also found that if the Reynolds number (i.e. velocity or mass)
of the approaching particles is low enough, penetration of the squeeze film on
rebound or even approach may not be possible, resulting in particle entrapment at the
target surface and a change in erosion mechanism. He observed that small particles
(less then100 micron) and concentrated slurries were especially liable to behave in
this way.
Gandhi B. K. et al15 [2004] have developed a methodology to determine the nominal
particle size of multi-sized particulate slurry for estimation of mass loss due to the
erosion wear. The effect of presence of finer particles (less then 75 micron) in
relatively coarse particulate slurry has also been studied. They have observed that
addition of particles finer than 75 micron in narrow-size or multi-sized slurries reduce
the erosion wear. In addition, the effective particle size for narrow-size particulate
slurries can be taken as the mean size whereas the weighted mass particle size seems
to be a better choice for multi-sized particulate slurries. The reductions in erosion
wear due to addition of fine particles decreases with increase in the concentration of
coarse size particles.
Rajesh J. John et al16 [2004] have carried out experiments on the effect of impinging
velocity on the erosive wear behaviour of polyamides. The impact angles were 30 and
60 degrees at two impact velocities (80 and 140 m/s). Silica sand is used as an
erodent. Surface blackening at the impact zone was observed for all the materials at
normal impact and at both the impact velocities. At normal impact and at lower
impinging velocity (i.e. 80 m/s), a mass gain in the initial period was observed for all
the materials except amorphous. The extent of increase in wear, however, depended
on the materials and the angle of impact. The velocity effect was more prominent at
the oblique angle of impact.
22
Wang Yao et al17 [2005] have designed an experimental system for wear assessment
of slurry pumps. Slurry density of 1450 kg/m3 has been taken. The wear patterns on
the suction liner, impeller and volute casing were shown. Two locations were found to
be most easily worn in impellers: the root of each vane and the area on the back
shroud close to the suction side of the vane root.
Tian Harry H. et al18 [2005] have observed the erosive wear of some metallic
materials such as high chromium white iron and aluminium alloy using Coriolis wear
testing approach. In the present study, the correlation between wear rate and particle
size on the tested materials is discussed. Factors, which should be considered in wear
modelling and prediction, have also been addressed. It can be seen that larger solids
particles resulted in higher mass loss in all test materials. Although the wear rates at
smaller particle sizes were relatively close within each material group, the wear rate
difference was significantly widened with larger particle sizes. The tested high-Cr
white irons showed a wear resistance some 27140 times higher than that of the
aluminium alloys in the Coriolis test conditions
concentration of slurry affected the wear results of the test materials. The higher the
flow rate, the higher the wear rate of test materials.
Asuaje Miguel et al19 [2005] have simulated the 3D-CFD model of the impeller and
volute of a centrifugal pump using CFX codes. A sensitivity analysis regarding grid
quality and turbulence models were also performed. A procedure for designing the
volute, the non-structured grid generation and the interface flow passage between the
impeller and volute were discussed. The flow simulations were carried out for several
impeller blades and volute tongue relative positions. It was observed that the
numerical solution is stabilized with grids containing more than 40 000 nodes. Also A
non uniform pressure distribution develops at the impeller periphery as a function of
the pump flow rate, causes a fluctuating radial thrust of significant amplitude.
Tian Harry H. et al20 [2005] have experimented on Coriolis wear testing. Wear
coefficients (or specific energy coefficients) have been determined for different slurry
conditions over a large range of particle sizes. Among the test materials, the harder
CrMo white iron alloy demonstrated the best wear resistance under slurry testing
conditions. It is also observed that Coriolis wear testing is an excellent approach to
simulate the erosive wear condition within a slurry pump. Beside particle sizes, other
23
particle properties such as particle shape and size distribution also exhibited
significant effect on the values of wear coefficients. Silica sand and copper ore
slurries were used as examples. The relationship between linear wear rate and solid
particles was also shown.
Das S K et al21 [2006] have developed a mathematical model embodying the
mechanisms of erosion on behaviour, to predict erosion rate of coal-fired boiler
components at different temperatures. Various grades of steels used in fabrication of
boiler components and published data pertaining to boiler fly ash have been used for
the modelling. The model incorporates high temperature tensile properties of the
target metal surface at room and elevated temperatures and has been implemented in
an user-interactive in-house computer code, to predict the erosion rates of various
grades of steel. The model is calibrated with plant and experimental data generated
from a high temperature air-jet erosion-testing facility
Li Ping et al22 [2006] have observed the failure analysis of the impeller used in zinc
hydrometallurgy process. Failure analysis revealed that an improper austenite/ferrite
ratio of duplex stainless steel (DSS) material resulting from too high nitrogen content
was primarily responsible for the rapid failure of the impeller. It was also observed
that nitrogen content had a significant effect on austenite/ferrite ratio of duplex steel.
The result of simulated corrosive wear test indicated that duplex stainless steel
possessing almost equal ferrite and austenite volume showed better corrosive wear
resistance.
Roudnev et al23 [2007] have observed the erosion wear of casing of slurry pumps and
predicted the accurate location of maximum erosion through CFD simulations and
experimental data. All 3D simulations predicted similar wear rate magnitudes. It is
being estimated that area nearer to upper part of volute casing experiences maximum
wear. CFD multiphase simulation using the Eulerian-Eulerian approach is being
applied to determine the velocity fields and solid concentration in the centrifugal
pump.
Shah Subhash N. et al24 [2007] have conducted experiments to visualize the effect of
slurry flow rate on the erosion of coiled tubing. Erosion rate has been investigated as a
function of flow rate, slurry concentration, solid particles size and density, and fluid
viscosity. This study utilizes both experimental tests and CFD simulations to
24
Feng Jianjun et al28 [2007] have conducted numerical simulations on impellerdiffuser interactions in radial diffuser pumps to investigate the unsteady flow, and
more attention is paid to pressure fluctuations on the blade and vane surfaces
Computational results show that a jet-wake flow structure is observed at the impeller
outlet. The biggest pressure fluctuation on the blade is found to occur at the impeller
trailing edge, on the pressure side near the impeller trailing edge, and at the diffuser
vane leading edge, independent of the flow rate, radial gap, and blade number
configuration. All of the flow rate, blade number configuration, and radial gap
influence significantly the pressure fluctuation and associated unsteady effects in the
diffuser pump.
26
27
CHAPTER 3
DESIGN OF CENTRIFUGAL PUMP
Before a pump can be selected or prototype designed, its application must be clearly
defined. As pump requirements must match system characteristics, analysis of overall
system is necessary in which the pump is to be installed .There are a lot of parameters
on which design of pump depends. For this design, head flow rate, speed and overall
efficiency has been taken as input.
The flow through the impeller is assumed to be resolved into two parts
a) Through flow with relative velocity assuming the impeller to be stationary
b) Rotational flow
The first part helps to design runner passages after assuming through flow velocity.
The second part helps to fix up outer diameter of impeller because this is mainly
responsible for developing the head required. The following procedure33 is followed
for design of centrifugal pump impeller.
3.1
INPUT DATA
Head (H)
3.2
14.5 m
Flow Rate ( Q ) =
0.016 m3/sec
Speed (N)
1450 rpm
DESIGN OF IMPELLER:
Ns
N Q
H0.75
(3.1)
28
Where
N = Speed at which pump shaft rotates=1450 rpm.
Q = Flow Rate =0.016 m3 / sec
H = Head =14.5 m.
Ns = 24.683 rpm
3.2.2. Input Power and Shaft Power
Output Power P0 =
gQH
(3.2)
gQH
)
745 0
(3.3)
Where
g=Acceleration due to gravity
=Density of the fluid
Overall efficiency is taken from the graph, and is taken equal to 0.75 percent. Input
power required 15% more because of bearing and transmission losses.
29
= 6.641 HP
Shaft power Psh = 1.15 Pin
Shaft Power Psh =
1.15 x 6.641
= 7.6373 HP
The relationship between hydraulic and overall efficiency33 is shown by eq. (3.4).
1 h 0.66 1 o
Hydraulic Efficiency,
h = 0.6436
Overall Efficiency,
o h v m
Mechanical Efficiency
m = 0.7768
(3.4)
(3.5)
Psh 60 750
) Nm
2 N
(3.6)
37472.5
Nm
dsh
16T 3
Fs
dsh
0.019689 m
(3.7)
Where
Fs
For the present problem the shaft is assumed to be made of steel with an
allowable shear stress of 25 MPa
30
3.2.4
3.2.5.
Dhb
1.2 x 0.019689
0.023627 m
(3.9)
Q' Q Q
(3.10)
Q'
0.016+0.016*0.08
Q'
0.01728 m3/sec
Where
Q
Q'
CA
2 tan 2 3 3
2 '
0.152
N Q
2
CA
2.5469 m/s
1
3
(3.11)
Where
CA
1.1
1.05
Constant
0.84
De
(3.12)
De
0.1014 m
D1
0.95 * De
0. 09633 m
1.05 * CA
2.6742 m/s
1.1 * C0
2.9416 m/s
C1
C0
D1
C0
C1
Where
b1
Q'
D1C1
0.02135 m
(3.13)
32
u1
D1 N
60
7.3135 m/s
C
tan 1 1
u1
21.91
(3.14)
2.6
(3.15)
The governing equation for solution to outlet end dimensions is the Eulers equation.
For rotation free inlet this may be written as:
Hr
u22 u2C2 m
cot 2
g
g
(3.16)
Where
Hr
C2m
u2
The left hand side of this equation is related to total head generated as given below.
Hr
H
h
1 p
(1+0.32)* 14.5/0.6436
29.7389 m
(3.17)
33
Where
p is Pfleiderers coefficient, a factor relating the difference between the imaginary and
ideal conditions inside the pump.
The value of meridional flow velocity inside an impeller does not vary much and can
be kept constant. However its value is generally somewhat lower than that at the
suction end and so assume
C2m
0.9 * C1m
0.9 * 2.9416
2.6474 m/sec
(3.18)
Where
C1m
u22 u2 2.6474
cot 260
9.81
9.81
20.0085 m/sec
D2
u 2 60
N
D2
0.2635 m
29.7389
Blade velocity at outlet
u2
Outer Diameter of impeller
(3.19)
34
D1 D2
sin 22
D2 D1
5.06
(3.20)
Q'
D 2 C2m
0.0078848 m
W22
C2
2 2
2g
2g
(3.21)
Take, 1 = 0.3,
2 =1.2
Where 1 is the constant considering losses in the impeller whose value is dependent
on the machine performance32, 2 is losses in suction pipe.
W2
C2
sin 2
(3.22)
8.8930
Where
W2 =
After substituting
1.6466 m
The results of the analytical work have been summarized in table 3.1
35
Design Parameter
Values
24.683
00.6436
37472.5 N-m
0.019689 m
Axial velocity
2.5469 m/s
Eye diameter
0. 09633 m
0.0963058 m
10
Hub diameter
0.0236283 m
11
Width at inlet
0.021342 m
12
7.3115 m/s
13
14
15
Number of blades
16
17
NPSH
0.776
7.6373 HP
20.0085 m/sec
0.2635 m
5
0.0078848 m
1.6466
36
CHAPTER 4
INTRODUCTION TO COMPUTATIONAL FLUID
DYANAMICS
______________________________________________
4.1 INTRODUCTION TO CFD
Computational fluid dynamics (CFD) is one of the branches of fluid mechanics that
uses numerical methods and algorithms to solve and analyze problems that involve
fluid flows. CFD has become an indispensable tool in the design, development,
evaluation and refinement of new industrial equipment and processes. The use of
CFD reduces the development cost of new products and cuts the time to market of
these products. CFD have applications in
1. Aerospace,
2. Automotive Design,
3. Electronics,
4. Food, Chemical, and Material Processing,
5. Power Generation
Computational Fluid Dynamics (CFD) has become an integral part of the engineering
design cycle. Whether it is used to design industrial mixing tanks, aerospace
components, or used to simulate transient flow phenomenon in pumps and turbines.
CFD analyses reduce development time and increase the reliability of prototype
designs. CFD simulations can be used to model fluid flows over a wide range of
physical scales. The fundamental basis of any CFD problem is the Navier-Stokes
equations, which define any single-phase fluid flow. It works by solving the equations
of fluid flow (in a special form) over a region of interest, with specified (known)
conditions on the boundary of that region.
37
38
.
Figure 4.1: Flow chart showing CFD solving methodology
The geometry of design needs to be created from the initial design. Any modelling
software can be used for modelling and then shifted to other simulation software for
analysis purposes.
4.3.3 Mesh generation
Mesh generation (Gridding) is the process of subdividing a region to be modelled into
a set of small control volumes. Associated with each control volume there will be one
or more values of the dependent flow variables (e.g., velocity, pressure, temperature,
etc.) Usually these represent some type of locally averaged values. Numerical
algorithms representing approximations to the conservation laws of mass, momentum
and energy are then used to compute these variables in each control volume.
39
Meshing is a method to define and break up the model into small elements. In general,
a finite element model is defined by a mesh network, which is made up of the
geometric arrangement of elements and nodes. Nodes represent points at which
features such as displacements are calculated. Elements are bounded by sets of nodes,
and define localized mass and stiffness properties of the model. Elements are also
defined by mesh numbers, which allow references to be made to corresponding
deflections, stresses, pressures, temperatures at specific model locations. The
traditional method of mesh generation is block-structure (multi-block) mesh
generation. The block-structure approach is simple and efficient technique of mesh
generation.
4.3.3.1 Topology
The topology is a structure of blocks that acts as a framework for positioning mesh
elements. Topology blocks represent sections of the mesh that contain a regular
pattern of hexahedral (hex) elements. They are laid out adjacent to each other without
overlap or gaps, with shared edges and corners between adjacent blocks, such that the
entire domain is filled. By using topology blocks to control the placement of hex
elements, a valid hexmesh can be generated to fill a domain of arbitrary shape. The
topology is invariant from hub to shroud and is viewed/edited on 2-D layers which are
located at various spanwise stations. The topology blocks can be arranged in a regular
(structured) pattern, an irregular (unstructured) pattern, or in a pattern consisting of
structured patches and unstructured patches. The choice of which approach should be
followed should be based on whichever method minimizes the maximum skew of the
topology blocks, since the skew in the hex elements of the mesh is directly related.
The topology should then be investigated at various layers (especially the hub and
shroud layers) to check its quality since the mesh quality is directly dependent on
topology. Topology blocks generally contain the same number of mesh elements
along each side. The mesh elements vary in size across topology blocks in a way that
produces a smooth transition within and between blocks. This is accomplished by
shifting the nodes toward, or away from, certain block edges.
There are basically 3 types of topologies which is generally used
1) H-Grid
2) J-Grid
40
3) O-Grid
In actual conditions, a combination of these topologies is used. An O-Grid is a
topology that forms a continuous loop around the blade profile. Using an O-Grid
around the blade yields excellent boundary layer resolution and near-orthogonal
elements on the blade.
Dominant
41
42
a) k- turbulence model
One of the most prominent turbulence models, the k (k-epsilon) model, has been
implemented in most general purpose CFD codes and is considered the industry
standard model. It has proven to be stable and numerically robust and has a well
established regime of predictive capability. For general purpose simulations, the
model offers a good compromise in terms of accuracy and robustness.
Within ANSYS CFX, the k turbulence model uses the scalable wall-function
approach to improve robustness and accuracy when the near-wall mesh is very fine.
The scalable wall functions allow solution on arbitrarily fine near wall grids, which is
a significant improvement over standard wall functions. While standard two-equation
models, provide good predictions for many flows of engineering interest k is the
turbulence kinetic energy and is defined as the variance of the fluctuations in velocity.
is the turbulence eddy dissipation (the rate at which the velocity fluctuations
dissipate), and has dimensions of per unit time. The k- model introduces two new
variables into the system of equations.
The continuity equation is then:
( U ) 0
t
(4.1)
(4.2)
where
B = Sum of body forces,
= Modified pressure
44
b) k- turbulence model
One of the advantages of the k formulation is the near wall treatment for lowReynolds number computations. The model does not involve the complex non-linear
damping functions required for the k model and is therefore more accurate and
more robust The models assumes that the turbulence viscosity is linked to the
turbulence kinetic energy and turbulent frequency via the relation:
k
w
(4.3)
Where
t = Turbulence viscosity
k = Turbulence kinetic energy and is defined as the variance of the
fluctuations in velocity
w=Turbulence frequency
c) Shear stress transport (SST) model
The Shear-Stress-Transport (SST) model is designed to give highly accurate
predictions of the onset and the amount of flow separation under adverse pressure
gradients by the inclusion of transport effects into the formulation of the eddyviscosity. This results in a major improvement in terms of flow separation predictions.
The superior performance of this model has been demonstrated in a large number of
validation studies 38. k model do not account for the transport of the turbulent
shear stress.The SST model is recommended for high accuracy boundary layer
simulations.The proper transport behavior can be obtained by a limiter to the
formulation of the eddy-viscosity
vt
a1k
max(a1, SF2 )
(4.4)
Where
vt t /
t = Turbulent viscosity
45
= Density
These models require the distance of a node to the nearest wall for performing the
blending between k and k turbulence models. Blending functions are used
for transformation of the one turbulence model to another
phase is present as
spherical particles of Mean Diameter d .Using this model, the inter phase contact
area is:
A 6r / d
(4.5)
Where
46
(4.6)
Where
superscripts o and n refer to old and new values respectively
v pi = Particle velocity.
47
t = Time Step
In forward integration, the particle velocity calculated at the start of the time step is
assumed to prevail over the entire step. At the end of the time step, the new particle
velocity is calculated using the analytical solution.
4.7)
In the calculation of all the forces, many fluid variables, such as density, viscosity and
velocity are needed at the position of the particle. These variables are obtained
accurately by calculating the element in which the particle is travelling, calculating
the computational position within the element.
4.3.4.4 Basic erosion model
a) Model of Finnie
The wear of a wall due to the erosive effect of particle impacts is a complex function
of particle impact, particle and wall properties. For nearly all metals, erosion is found
to vary with impact angle and velocity according to the relationship
E kVPn f ( )
(4.8)
Where
E = dimensionless mass
f ( ) 1/ 3cos2
if tan 1/ 3
(4.9)
f ( ) sin(2 ) 3sin 2
if tan 1 / 3
(4.10)
48
/ 2
f ( ) 1 k 2k12 sin
(4.11)
(4.12)
R T 1 k 4 VP sin
(4.13)
(4.14)
1.0 if 2 0
0.0 if 2 0
(4.15)
Where
E= Dimensionless mass (mass of eroded wall material divided by the mass
of particle)
Vp= Particle impact velocity
= Impact angle in radians between the approaching particle track and the
wall,
49
velocity coefficients of restitution depend upon the hardness of the target material, the
density of the particle and the impact velocity.
4.3.5 Solver
Navier-Stokes Equation
The set of equations which describe the processes of momentum, heat and mass
transfer are known as the Navier-Stokes equations. These partial differential equations
were derived in the early nineteenth century and have no known general analytical
solution but can be discretized and solved numerically. Equations describing other
processes, such as combustion, can also be solved in conjunction with the NavierStokes equations. Often, an approximating model is used to derive these additional
equations, turbulence models being a particularly important example. There are a
number of different solution methods which are used in CFD codes. The most
common, is known as the finite volume technique. In this technique, the region of
interest is divided into small sub-regions, called control volumes. The equations are
discretized and solved iteratively for each control volume. As a result, an
approximation of the value of each variable at specific points throughout the domain
can be obtained. In this way, one derives a full picture of the behaviour of the flow.
The Navier-Stokes equations, named after Claude-Louis Navier and George Gabriel
Stokes, describe the motion of fluid substances such as liquids and gases. These
equations establish that changes in momentum in infinitesimal volumes of fluid are
simply the sum of dissipative viscous forces (similar to friction), changes in pressure,
gravity, and other forces acting inside the fluid: an application of Newton's second
law.
They are one of the most useful sets of equations because they describe the physics of
a large number of phenomena of academic and economic interest. They may be used
to model weather, ocean currents, water flow in a pipe, flow around an airfoil (wing),
and motion of stars inside a galaxy. As such, these equations in both full and
simplified forms, are used in the design of aircraft and cars, the study of blood flow,
the design of power stations, the analysis of the effects of pollution, etc. Coupled with
Maxwell's equations they can be used to model and study magneto hydrodynamics.
50
X-Momentum
2 u 2 u
u
u
u
u
v 2 2
t
x
y
y x
x
(4.16a)
Y-Momentum
2v 2v
v
v
v
u v 2 2
t
x
y
x y x
(4.16b)
And continuity
u v
0
x y
(4.17)
Where
u = Component of Velocity in x-direction
v = Component of Velocity in y-direction
= Viscosity of Fluid
= Density of fluid
51
52
CHAPTER 5
MODELLING AND ANALYSIS
___________________________________________________
5.1 IMPELLER 3D FLOW SIMULATIONS
A 3D flow simulation is carried out on an isolated impeller of a radial flow centrifugal
pump on ANSYS-CFX.
CFX. The procedure adopted for solving the given problem is
described in the previous chapter
5.2 INITIAL DESIGN
A design code has been developed using visual C++ for calculation of design of
centrifugal pump and verified with the analytical work done in the chapter 3
5.3 GEOMETRY GENERATION
54
5.5 PREPROCESSING
5.5.1 Simulation Parameters and Boundary Conditions
The general parameters and boundary conditions used for 3D flow simulations of the
impeller are summarized in the table 5.1.
For all simulations the boundary conditions are as follows:
1) Inlet: total pressure applied in the rotation axis direction
2) Outlet: imposed mass flow
3) Periodic: two symmetry surfaces positioned in the middle of the blade
passage.
4) Wall: general boundary conditions by default.
The simulation domain at the inlet and outlet sections is sufficiently extended to allow
recirculation and the elliptic influence of the flow. Table 5.1 give details of impeller
simulation parameters which have been defined while defining boundary conditions.
55
Grid
Structured
Fluid
Turbulence model
SST model
Discretization
Second order
10 4
conditions.
However Shear stress transport model is found to be more suitable for turbo
machinery problems as they have features of both the k , k models. So SST
model is chosen for the present work.
56
Variable
Coefficient
Value
k12
k12
0.293328
Ref Velocity 1
V1
123.72 [m/s]
Ref Velocity 2
V2
352.99 [m/s]
Ref Velocity 3
V3
179.29 [m/s]
Angle of Maximum
30
Erosion
57
For all cases except changing particle size common sand is taken with sizes and
percentage given below in Table 5.3
Table 5.3 Sand particle size and percentage distribution
No.
Sand sizes(microns)
Percentage distribution
50
0.02
100
0.18
200
0.2
300
0.4
600
0.2
5.6 SOLVER
There are several methods of discretizing a given differential equation, but finite
volume is used in ANSYS-CFX. The finite volume method is a numerical method for
solving partial differential equations that calculates the values of the conserved
variables averaged across the volume. One advantage of the finite volume method
over finite difference methods is that it does not require a structured mesh (although a
structured mesh can also be used). Furthermore, the finite volume method is
preferable to other methods as a result of the fact that boundary conditions can be
applied noninvasively. This is true because the values of the conserved variables are
located within the volume element, and not at nodes or surfaces. Finite volume
methods are especially powerful on coarse non uniform grids and in calculations
where the mesh moves to track interfaces or shocks.
58
CHAPTER 6
RESULTS AND DISCUSSIONS
___________________________________________________
6 .1 DESIGN OF IMPELLER
The design data can be verified from the analytical work done in the chapter 3 and
summarized in table 3.1
1) Specific speed of the pump is:
24.6832
0.6436
0.77688
7.63732 HP
5689.8 Watts
6) Torque:
37472.5 N-m
0.0196902 meters
8) Axial velocity:
2.54876
9) Eye diameter:
0.101375 meters
0.0963058 meters
0.0236283 meters
0.021342 meters
21.9318 degree
0.270368 meters
59
0.00767887 meters
1.6492
26 degree
60
0.0000007
0.0000006
0.0000005
0.0000004
Erosion Rate
from inlet to
outlet
0.0000003
0.0000002
0.0000001
0
0
20
40
60
however the erosion rate at the trailing edge is less severe. The highest value of
erosion rate density at inlet is three times more than erosion rate density at trailing
edge .It signifies that wear pattern at inlet is more severe than wear at the trailing edge
of the blade This has been visualized on Figure 6.4.
. Figure 6.4 Experimental results shows complete wear of blades near the inlet22
62
Erosion Rate
of blade at 5%
concentration
4.00E-06
3.50E-06
3.00E-06
2.50E-06
Erosion Rate
of blade at
10%
concentration
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
0
200
400
600
800
1000
1200
1400
1600
Erosion Rate
of blade at
15%
concentration
Particle Size(Microns)
Figure 6.6 Erosion rate vs. particle size for blade at different concentrations
The erosion rate density of the blade is nearly 8-10 times greater than hub, which
indicates that the hub is less prone to wear. The simulation is conducted at 5%, 10%
and 15% and a similar trend is observed. The erosion rate trend of hub is totally
different than that of the blade. It shows nearly equal erosion rate trend at lower
values of particle size but increases to higher values above 1000 microns particle size.
It clearly indicates that the manner in which erosion occurs in the blade and the hub
63
depends upon the manner in which particles strikes the surface. However, the graphs
clearly justify the fact that blades of the impeller are more susceptible to erosion.
0.0000008
0.0000007
Erosion Rate
of hub at 5%
concentration
0.0000006
0.0000005
0.0000004
Erosion Rate
of hub at 10%
concentration
0.0000003
0.0000002
0.0000001
0
-1E-07 100
300
500
700
900
1100
1300
1500
Erosion Rate
of hub at 15%
concentration
Particle SIze(Microns)
Erosion Rate of
blade at 5%
concentration
4.00E-06
Erosion Rate (kg/s/m^2))
3.50E-06
Erosion Rate of
blade at 10%
concentration
3.00E-06
2.50E-06
Erosion Rate of
hub at 5%
concentration
2.00E-06
1.50E-06
Erosion Rate of
hub at 10%
concentration
1.00E-06
5.00E-07
0.00E+00
-5.00E-07 100
300
500
700
900
1100
Particle size(microns)
1300
1500
Erosion Rate of
blade at 15%
concentration
Erosion Rate of
hub at 15%
concentration
Figure 6.8 Erosion rate predictions on blade and hub at different concentrations
and varying particle size
64
1.40E-06
1.20E-06
Erosion Rate at
blade at 5%
concentration
1.00E-06
8.00E-07
Erosion Rate at
blade at 10 %
cencentration
6.00E-07
4.00E-07
Erosion Rate at
blade at 10 %
cencentration
2.00E-07
0.00E+00
8
10
12
14
16
18
20
Figure 6.9 Erosion rate vs. flow rate at different concentrations on blade
The pattern of increase in erosion rate is nearly same for 5%, 10% and 15%
concentration. And there is a linear increase in erosion rate with flow rate. At low
flow rates the effects are negligible but as the flow rate increases, the erosion rate
increases.
2.50E-06
Erosion Rate
Density of
blade at 5 %
concentration
2.00E-06
1.50E-06
Erosion Rate
Density of
blade at 10 %
concentration
1.00E-06
5.00E-07
0.00E+00
1000
1200
1400
1600
1800
2000
Erosion Rate
Density of
blade at 15 %
concentration
RPM (Degrees)
Erosion Rate
Density of hub
at 5 %
concentration
Erosion Rate
Density of hub
10 %
concentration
1200
1400
1600
1800
2000
Erosion Rate
Density of hub
15 %
concentration
RPM (Degrees)
66
Erosion Rate(kg/s/m^2)
2.50E-06
2.00E-06
1.50E-06
1.00E-06
5.00E-07
0.00E+00
1000
1200
1400
1600
1800
2000
SPEED (Degrees)
Figure 6.12 Erosion rate vs. Speed on blade and hub at different concentrations
Ersion Rate(kg/s/m^2)
3.00E-06
2.50E-06
Erosion
rate
(blade)
2.00E-06
1.50E-06
Erosion
Rate (hub)
1.00E-06
5.00E-07
0.00E+00
5
10
15
20
25
30
Concentration (%)
67
kg
m 2 s 1000
sm 2
68
69
70
Velocity also increases gradually along streamwise direction within the impeller
passage. As the flow enters the impeller eye, it gets diverted to the blade-to-blade
passage. The flow at the entrance is not smooth because of the unsteady flow entering
the impeller passage. The separation of flow can be seen at the blade leading edge.
Since, the flow at the inlet of impeller is not tangential to the blade, the flow along the
blade is not uniform and hence the separation of flow takes place along the surface of
the blade. Figure 6.18 shows the velocity distribution from inlet to outlet in a
centrifugal pump impeller.
71
CHAPTER 7
CONCLUSIONS AND SCOPE FOR FUTURE WORK
___________________________________________________
7.1 CONCLUSIONS
1) The results show that the erosion wear has strong dependence on flow rate. The
higher the flow rate, the higher the wear rate of the test material. Solid concentration
also affects the wear rate linearly.
2) The effect of erosion rate is more on the blades than on the hub. The graph shows
that the erosion rate density on the blades is nearly 8-10 times than on the hub.
3) The effect of wear is more at leading edge than at trailing edge of blades. So
selecting the optimized inlet blade angle can play a major role in reducing erosion
wear and providing smooth flow for fluid.
4) The effect of particle size on the erosion rate is predominant only up to certain size
limit i.e. 600-800 microns on the blades. After that, the erosion rate increases with a
comparatively slower rate. This indicates that the particle size affects the erosion rate
only up to certain particle size.
5) Increasing flow rate causes sharp decrease in pressure which can cause cavitation
on the surface of blade.
While designing the components handling slurry, it is necessary to select proper
material to increase their service life. Different methods can be used to enhance the
surface properties to decrease the erosion wear and thus to increase the service life of
components. Heat treatment is the best method to increase the surface hardness.
Martensite structure (which has the highest hardness) can be developed on the surface
of the component, thus retaining higher toughness from inside. The experiments have
been conducted on the steels and it is found that the erosion rate decrease with
increasing hardness by heat treatment of the steels.
Ceramic coating has been suggested4 to protect the pump due to slurry erosion.
Instead of applying coating to whole impeller, the coating can be applied only to
72
blade to cut manufacturing cost. Sharp corners should be avoided; the root of the
blade should be round to provide smooth flow of fluid.
73
REFERENCES
___________________________________________________
1. Desai P. V ,Pagalthivarthi V, Addie G. R. (1990), Particular motion and
concentration fields in centrifugal slurry pumps, Particulate Science and Technology,
Vol. 8:1, pp 77 96
2. Achim Daniela,Alan K Easton,Phillip M. Schwarz,Peter J Witt and Zakhari (1999),
Computational and experimental studies of tube erosion in a fluidised bed , Second
International Conference On CFD in the Minerals and Process Industries,CSIRO,6-8
December, pp 243-248
3..Manickam M., M.P. and Schwarz, Mcintosh M.J,(1999), CFD analysis of erosion
of bifurcation duct wall ,Second International Conference On CFD in the Minerals
and Process Industries,CSIRO,6-8 December, pp 243-248
4. Zhao H.K, Yabuki A,Matsumura M ,Takahashi,T and Yamamoto M,(1999),
Slurry erosion properties of ceramic coating, Wear,Vol. 233-235, pp 608-614.
5. Anders Sellgren, Addie Graeme and Yu Wi-Chung, (1999), Effects of nonnewtonian mineral suspensions on the performance of centrifugal pumps, Mineral
Processing and Extractive Metallurgy Review, Vol20:1, pp 239 249
6. Lee S. Y., Dimenna R. A., and Duignan M. R. 2001, Designing a scaled erosion
test with CFD methods, U.S. Department of Energy WSRC-MS-2001-00500
7. Gandhi B.K, Singh S. N., Seshadri V. 2001, Performance characteristics of
centrifugal slurry pumps, Journal of Fluids Engineering ,Vol. 123, pp 271-280
8. Hawthorne H.M. 2002, Some Coriolis slurry erosion test developments,
Tribology International, Vol. 35 ,pp 625630
9. Bakir F., Kouidri S., Belamri T., and Rey R.,(2001), On a general method of
unsteady potential calculation applied to the compression stages of a turbo machine
Part I: Theoretical approach, Journal of Fluids Engineering, Vol. 123, no. 4, pp. 780
786.
74
10. Clark H. McI. , (2002), Particle velocity and size effects in laboratory slurry
erosion measurements or do you know what your particles are doing, Tribology
International, Vol. 35, pp 617624
11. Zhou Weidong, Zhao Zhimei, Lee T. S. and Winoto S. H,(2003), Investigation of
flow through centrifugal pump impellers using computational fluid dynamics,
International Journal of Rotating Machinery, Vol. 9,Issue 1, pp 49-61
12.. Hamed Awatef A., Tabakoff Widen, Rivir Richard B., Das Kaushik, Arora
Puneet,(2002), Turbine blade surface deterioration by erosion, Journal of
Turbomachinery, July Vol 127, Issue 3, pp. 445-452
13. Engin T, Gur M, 2003, comparative evaluation of some existing correlations to
predict head degradation of centrifugal slurry pumps, Journal of Fluids Engineering,
Vol. 125, pp 149-157
14. Clark H.McI. 2004, The influence of the squeeze film in slurry erosion, Journal
of Wear, Vol. 256 , pp 918926
15. Gandhi B K., Borse Satish V. 2004, Nominal particle size of multi-sized
particulate slurries for evaluation of erosion wear and effect of fine particles, Journal
of Wear, Vol. 257 , pp 73-79
16. Rajesh J. John, Bijwe J., Venkataraman B., Tewari U.S. 2004, Effect of
impinging velocity on the erosive wear behaviour of polyamides, Tribology
International Vol. 37 , pp 219-226
17. Wang Yao, Zuo Ming J., and Fan Xianfeng 2005, Design of an experimental
system for wear asssessment of slurry pumps, Proceeding of the 2nd CDENconference 2005, Kananaskis, University of Calgary
18 Tian Harry H., Graeme R. Addie 2005, Experimental study on erosive wear of
some metallic materials using Coriolis wear testing approach, Journal of Wear vol.
258 , pp 458469
19. Asuaje Miguel, Bakir Farid, Kouidri Smaine, Kenyery Frank, Rey Robert 2005,
Numerical modelization of the flow in centrifugal pump: Volute Influence in
75
76
BIBLIOGRAPHY
29. A.J Stepanoff,(1958), Centrifugal and axial flow pumps theory design and
application, 2nd ed, John wilay and sons, New York
30. A .J Stepanoff,(1965), Pumps and blowers and two-phase flow, John wilay and
sons, New York
31. John Evans Miller, Frederick Schmid,(1987), Slurry erosion: uses, applications,
and test methods: a Symposium, Astm International
32. Labanoff Val S,(1992), Centrifugal pumps design and applications, 2nd edition
Gulf Publishing Company.
33. Vasandani V.P., (1993), Hydraulic machines theory and design, 10th edition,
Khanna Publishing house, New Delhi.
34. P.S Ghoshdastidar (1998), Computer simulation of flow and heat transfer, 1998,
Tata McGraw-Hill.
35. John Tuzson, (2000),Centrifugal pump design, September 2000, Wiley
publisher
36. Igor J Karassik,(2001), Pump handbook, Third Edition, McGraw-Hill
Professional
37 Robert Lafore, (2005), Object oriented programming in turbo C++, Galgotia
publications pvt ltd
38. ANSYS CFX-Manual,(2006),Published by, ANSYS CFX, Release 11.0,
December, 2006, ANSYS, Inc.
77
ANNEXURE I
PROGRAM FOR DESIGN OF CENTRIFUGAL PUMP
___________________________________________________
#include<stdafx.h>
#include<iostream>
using namespace std;
double a,b,c,u1;
void fn_Roots()
{
double D = (b*b-4*a*c);
if(D>=0)
{
double x1 = (-b+sqrt(D))/(2*a);
double x2 = (-b-sqrt(D))/(2*a);
if(x1>=0)
{
u1=x1;
}
if(x2>=0)
{
u1=x2;
}
}
if(D<0)
cout<<"\nEquation has imaginary roots"<<endl;
}
78
void main()
{
//Enter the input parameters//
double H,Q,p,N,Ns,Pin,P;
double Oeff,Pw,Heff,Meff,Veff,pi=3.1415;
double theta=3.143*22/180,g=9.81,C1m,beta1;
double
Q1,De,Ca,B1,b1,b2,a1,a2,a3,Beta2,C2,C3,u2,D1,Dh,Hr,C2m,w,dH,beta;
double T,Dsh,D2,D,b0,Ts,Cl,G1;
double phi,sai,Hs,w1,s1,r1;
double R,R1,dr=0.002;
int
z;
//Density of water
79
cout<<"********************************************************
********************"<<endl;
cout<<"PROGRAM FOR THE INPUT PARAMETERS FOR THE DESIGN
OF CENTRIFUGAL PUMP IMPELLER"<<endl;
cout<<"********************************************************
********************";
cout<<"\nEnter the value of desired head of the pump (meters): "<<"\t";
cin>>H;
cout<<"\nEnter the value of desired flow rate (m^3 per sec): "<<"\t";
cin>>Q;
cout<<"\nEnter the value of desired speed of the pump (rpm): "<<"\t";
cin>>N;
cout<<"\nEnter the value of overall efficiency of pump (%): "<<"\t";
//Take the value from the Stepanoff's curves
cin>>Oeff;
cout<<"\nEnter the value of Vol. efficiency of the pump(%): "<<"\t";
//Normally varies from 0.85 to 0.98
cin>>Veff;
cout<<"********************************************************
********************";
cout<<"\nCONSTANTS USED IN THE CALCULATION \n"
cout<<"********************************************************
*** *****************";
cout<<"\n\nDensity of water(r):"<<"\t\t\t\t\t"<<r<<" kg/cubic metre";
cout<<"\nAccleration due to gravity(g):"<<"\t\t\t\t"<<g<<" metre per sec sqr";
cout<<"\nShear stress of the shaft in MPa(s):"<<"\t\t\t"<<s;
80
cout<<"\nsigma:"<<"\t\t\t\t\t\t\t"<<sigma;
cout<<"\ntow:"<<"\t\t\t\t\t\t\t"<<tow;
cout<<"\nk:"<<"\t\t\t\t\t\t\t"<<k;
cout<<"\n*******************************************************
*********************";
cout<<"\nRESULTS OF THE GIVEN INPUT"<<"\n";
cout<<"\n*******************************************************
*********************";
p=N*sqrt(Q);
Ns=p/pow(H,0.75);
hydraulic efficiency
cout<<"\nHydrualic efficiency of the pump :"<<"\t\t\t"<<Heff<<"%"<<"\n\n";
Meff=(Oeff)/(Heff*Veff);
cout<<"\nMechanical Eff. of the pump: "<<"\t\t\t"<<Meff<<"%"<<"\n\n";
P=(r*g*Q*H)/(745*Oeff);
Pin=1.15*P;
Pw=Pin*745;
cout<<"\nInput power of the pump: "<<"\t\t\t\t"<<Pin<<" HP "<<endl;
cout<<"\nInput power of the pump: "<<"\t\t\t\t"<<Pw<<" Watts "<<endl;
//Calculation of Shaft diameter
81
T=Pw*60*1000/(2*pi*N);
cout<<"\nTorque: "<<"\t\t\t\t\t\t"<<T<<endl;
D=16*T/(pi*s);
Dsh=pow(D,0.333333);
cout<<"\nDiameter of the Shaft: "<<"\t\t\t\t"<<Dsh<<" mm"<<endl;
Q1=Q+(Q*0.08);
a1=sigma*sigma*tan(beta3)*tan(beta3);
a2=tow*tow*k;
a3=a1/a2;
a=pow (a3,0.3333);
b1=N*N*Q1;
b=pow (b1,0.3333);
Ca=0.152*a*b;
//Axial velocity
//Dia of eye
D1 = (60*u1)/(pi * N);
//CALCULATION OF NUMBER OF BLADES
cout<<"\nOutlet diameter of blade: "<<"\t\t\t\t"<<D1<<"
meters"<<endl;
z=2*pi*sin(theta)*((D2+D1)/(D1-D2));
cout<<"\nNumber of blades: "<<"\t\t\t\t\t"<<z+1<<endl;
B1=(Q1)/(pi*D1*C1m);
cout<<"\nBreadth at outlet of blade"<<"\t\t\t\t"<<B1<<"
meters"<<endl;
double t1=0.3,t2=1.2;
//
CALCULATION OF NPSH
w=C3/sin(beta3);
dH=((t1*w*w)+(t2*C3*C3))/(2*g);
cout<<"\nNPSH Required is:"<<"\t\t\t\t\t"<<dH<<endl;
cout<<"\nOutlet blade angle: "<<"\t\t\t\t\t"<<outang<<"
degree"<<"\n\n";
cout<<"**************************************************
**************************";
84
ANNEXURE II
OUTPUT OF THE PROGRAM
*********************************************************************
PROGRAM FOR THE INPUT PARAMETERS FOR THE DESIGN OF
CENTRIFUGAL PUMP IMPELLER
*********************************************************************
ENTER THE VALUE OF DESIRED HEAD OF THE PUMP (METERS): 14.5
ENTER THE VALUE OF DESIRED FLOW RATE (M^3 PER SEC):
0.016
1450
0.46
0.92
*********************************************************************
CONSTANTS USED IN THE CALCULATION
*********************************************************************
DENSITY OF WATER(R):
ACCLERATION DUE TO GRAVITY (G):
SQR
SHEAR STRESS OF THE SHAFT IN MPA(S):
25
SIGMA:
1.1
TOW:
1.05
K:
0.84
********************************************************************
85
24.6832
0.6436%
0.77688%
7.63732 HP
5689.8 WATTS
TORQUE:
37472.5
19.6902 MM
AXIAL VELOCITY:
2.54876
EYE DIAMETER:
0.101375 METERS
0.0963058 METERS
HUB DIAMETER:
23.6283 MM
WIDTH AT INLET:
0.021342 METERS
21.9318 DEGREE
0.270368 METERS
NUMBER OF BLADES:
0.00767887 METERS
1.6492
26 DEGREE
*********************************************************************
86